Multilingual user feedback analysis system and related methods

Abstract

The system (102) for analysing user feedback includes a memory (206) and a processor (202). The processor (202) executes programmed instructions for receiving user interactions in multiple languages, processing these interactions to extract information and sentiment, and identify relevant entities. It also generates condensed information of the processed user interactions at multiple levels and continuously learns on the processed user interactions to enhance its accuracy in producing condensed information over time. The system also outputs the generated multilevel condensed information to inform product improvements and feature prioritization decisions.

Description

[0001] MULTILINGUAL USER FEEDBACK ANALYSIS SYSTEM AND RELATED

[0002] METHODS

[0003] CROSS-REFERENCE TO RELATED APPLICATIONS AND PRIORITY

[0004] The present application claims priority from the Indian patent application having application number 202411033569 filed on 27 April 2024, incorporate herein by a reference.

[0005] FIELD OF INVENTION

[0006] The present invention, in general, relates to the field of natural language processing and machine learning and more particularly, relates to multilingual user feedback analysis system and related methods.

[0007] BACKGROUND

[0008] In the era of digitalization, user feedback plays a pivotal role in the development and improvement of products and services. Users across the globe interact with digital platforms in various languages, expressing their thoughts, concerns, and suggestions. This feedback is a goldmine of information that can guide product development, feature enhancement, and user experience improvement. However, the sheer volume and diversity of this feedback present significant challenges. Manual analysis is time-consuming and prone to errors, while traditional automated methods may lack the ability to accurately capture the nuances of user sentiment, especially when dealing with multiple languages. Furthermore, the dynamic nature of language, with evolving slang, informal expressions, and new terminology, adds another layer of complexity to the analysis. Therefore, there is a need for an efficient and effective method to process and analyze multilingual user feedback.

[0009] Thus, there is a long-felt need for multilingual user feedback analysis system and related methods.

[0010] SUMMARY

[0011] This summary is provided to introduce concepts related to multilingual user feedback analysis system and related methods, and the concepts are further described below in the detailed description. This summary is not intended to identify essential features of the claimed subject matter nor is it intended for use in determining or limiting the scope of the claimed subject matter.

[0012] In accordance with embodiments, a system and method are provided for analysing user feedback. The system comprises a memory and a processor configured to execute programmed instructions stored in the memory. The system receives user interactions in a plurality of languages, processes these interactions using a natural language processing (NLP) module to extract essential information and sentiment, and identifies relevant entities. The system generates condensed information of the processed user interactions at multiple levels using a machine learning (ML) summarization module. This module continuously learns on the processed user interactions, adjusting to changing patterns and expressions to enhance its accuracy in producing condensed information over time. The system outputs the generated multilevel condensed information to inform product improvements and feature prioritization decisions.

[0013] In accordance with other embodiments, a method for analysing user feedback is provided. The method involves receiving user interactions in a plurality of languages, processing these interactions using a natural language processing (NLP) module to extract essential information and sentiment, and identifying relevant entities. The method generates condensed information of the processed user interactions at multiple levels using a machine learning (ML) summarization module. This module continuously learns on the processed user interactions, adjusting to changing patterns and expressions to enhance its accuracy in producing condensed information over time. The method outputs the generated multilevel condensed information to inform product improvements and feature prioritization decisions.

[0014] BRIEF DESCRIPTION OF DRAWINGS

[0015] The detailed description is described with reference to the accompanying Figures. The same numbers are used throughout the drawings to refer like features and components.

[0016] Figure 1 illustrates a network implementation of a system, in accordance with an embodiment of the present disclosure;

[0017] Figure 2 illustrates a block diagram of the system, in accordance with an embodiment of the present disclosure;

[0018] Figure 3 illustrates a flowchart for multilingual user feedback analysis; and

[0019] Figure 4 illustrates a flowchart depicting a method of multilingual user feedback analysis, in accordance with an embodiment of the present disclosure. DETAILED DESCRIPTION

[0020] Reference throughout the specification to “various embodiments,” “some embodiments,” “one embodiment,” or “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment. Thus, appearances of the phrases “in various embodiments,” “in some embodiments,” “in one embodiment,” or “in an embodiment” in places throughout the specification are not necessarily all referring to the same embodiment. Furthermore, the particular features, structures or characteristics may be combined in any suitable manner in one or more embodiments.

[0021] Referring to figure 1, network implementation 100 of a system 102 for tiered withdrawal system for optimizing user engagement and mitigating operational expenses is illustrated, in accordance with an embodiment of the present subject matter. The online platform may be a gaming platform, a social media platform, and the like. In one embodiment, the system 102 may comprise a processor and a memory. Further, the system 102 may be connected to user devices 104 or applications residing over the user devices 104 through a network 106. It may be understood that the system 102 may be communicatively coupled with the user through one or more user devices / applications 104-1, 104-2, . . ., 104-n collectively referred to as a user device 104.

[0022] In one embodiment, the network 106 may be a cellular communication network used by user devices 104 such as mobile phones, tablets, or a virtual device. In one embodiment, the cellular communication network may be the Internet. The user device 104 may be any electronic device, communication device, image capturing device, machine, software, automated computer program, a robot or a combination thereof. The system 102 may be configured to register users over the system 102.

[0023] In one embodiment, the user devices 104 may support communication over one or more types of networks in accordance with the described embodiments. For example, some user devices and networks may support communications over a Wide Area Network (WAN), the Internet, a telephone network (e.g., analog, digital, POTS, PSTN, ISDN, xDSL), a mobile telephone network (e.g., CDMA, GSM, NDAC, TDMA, E-TDMA, NAMPS, WCDMA, CDMA-2000, UMTS, 3G, 4G), a radio network, a television network, a cable network, an optical network (e.g., PON), a satellite network (e.g., VSAT), a packet-switched network, a circuit- switched network, a public network, a private network, and / or other wired or wireless communications network configured to carry data. The aforementioned user devices 104 and network 106 may support wireless local area network (WLAN) and / or wireless metropolitan area network (WMAN) data communications functionality in accordance with Institute of Electrical and Electronics Engineers (IEEE) standards, protocols, and variants such as IEEE 802.11 (“WiFi”), IEEE 802.16 (“WiMAX”), IEEE 802.20x (“Mobile-Fi”), and others. The block diagram of the of the system 102 is further illustrated in figure 2.

[0024] Referring now to figure 2, various components of the system 102 are illustrated, in accordance with an embodiment of the present subject matter. As shown, the system 102 may include at least one processor 202, an I / O interface 204 and a memory 206. The memory 206 consists of programmed instructions corresponding to a set of modules 208 and data 210. The set of modules 208 may include an input module 212, a natural language processing (NLP) module 214, a machine learning (ML) summarization module 216, and an output module 218. In one embodiment, the at least one processor 202 is configured to fetch and execute computer-readable instructions, stored in the memory 204, corresponding to each module 208. It must be noted that though the invention is explained considering that the system 102 is deployed over a remote server and the user device 104 is communicatively coupled with the system 102 through the network 106. However, it must be noted that the system 102 or modules 208 of the system 102 may also be deployed on the user device 104 itself and the modules 208 perform the same functions as that on the server using the local hardware of the user device 104. By implementing the system 102 on the user device 104, the data privacy of user’s personal data is maintained as the data never leaves the user device 104.

[0025] In one embodiment, the memory 204 may include any computer-readable medium known in the art including, for example, volatile memory, such as static random-access memory (SRAM) and dynamic random-access memory (DRAM), and / or non-volatile memory, such as read-only memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and memory cards.

[0026] In one embodiment, the programmed instructions may include routines, programs, objects, components, data structures, etc., which perform particular tasks, functions, or implement particular abstract data types. The data 210 may comprise a data repository 222, and other data 224. The other data 224 amongst other things, serves as a repository for storing data processed, received, and generated by one or more components and programmed instructions. The working of the system 102 will now be described in detail referring to figure 2 and figure 3. In one embodiment, the processer 202 may execute programmed instructions corresponding to the input module 212, the natural language processing (NLP) module 214, the machine learning (ML) summarization module (hereinafter ML module) 216, and the output module 218. The multilingual user feedback analysis system 102 may be configured to interpret and summarize user feedback across languages, aiding in product development and decisionmaking. The system includes a processor 202 and memory 206, which execute instructions for analysing feedback. The input module 212 collects user interactions, while the NLP module (hereinafter NLP module) 214 processes this data. The ML module 216 generates and refines summaries, and the output module 218 disseminates insights.

[0027] The system 102 initiates its process with the input module 212 receiving user interactions in multiple languages. This data acquisition ensures that the system's analysis includes a wide range of user inputs. The NLP module 214 then processes the text, breaking it down into individual words or phrases, and gaining insights around product features or bugs.

[0028] The ML module 216 takes the processed data and condenses it into summaries. The ML module 216 identifies informative sentences and generate embeddings out of them. The ML Module 216 also groups these summaries by themes or features, which aids in organizing the feedback for review. To ensure the system remains effective, the ML module 216 may be trained on a dataset of previously processed feedback, allowing it to learn historical language patterns and expressions. This continuous learning process may be integrated into the system's operation. The output module 218 then presents the multilevel summaries in a structured format, which includes individual, feature-wise, daily, and monthly insights. These summaries provide product managers with the necessary information to make informed decisions regarding product improvements, bug resolution, and feature prioritization.

[0029] The processor 202 may be a central element in the multilingual user feedback analysis system, orchestrating the execution of programmed instructions to analyze user feedback across languages. The processor 202 executes instructions for receiving user interactions, processing them via natural language processing, generating summaries with a machine learning module, and outputting insights. The processor 202 initiates the input module 212 to collect user interactions in various languages. This step ensures the system has access to a wide range of user feedback for analysis. Following data collection, the processor 202 utilizes the NLP module 214 to dissect the textual content. This process includes tokenization, which breaks down the text into individual words or phrases, and part-of- speech tagging. Additionally, named entity recognition may be used to identify and classify significant elements within the text.

[0030] The processed data may be then directed to the ML module 216, which synthesizes the information into summaries. This module selects informative sentences and embeds them to retain essential information, organizing them by themes or features through topic modelling. The ML module 216 may be designed to learn from new data, enhancing its summarization capabilities over time.

[0031] Finally, the processor 202 commands the output module 218 to format and present the multilevel summaries. These summaries provide a structured breakdown of individual interactions, feature-specific insights, daily trends, and long-term analyses, which support product managers in making decisions regarding product development and feature prioritization.

[0032] The NLP module 214 serves as a key functional unit within the system, tasked with interpreting user feedback. It operates by converting human language into a format that can be understood and processed by machines, which may be essential for extracting meaningful data from user inputs in various languages.

[0033] The NLP module 214 comprises several processes, starting with tokenization, which segments text into individual units such as words or phrases. This process may be the initial step in preparing the data for detailed analysis and may be necessary for identifying linguistic patterns within the text. This process may be vital for revealing the structure of sentences and contributes to the understanding of syntactic meaning within the user feedback. The final process within the NLP module 214 may be named entity recognition, which locates and categorizes specific entities within the text, such as product features or user locations. This process may be dependent on the successful completion of the previous steps, as it requires the text to be tokenized to accurately classify entities. The combined output of these processes equips the system with a comprehensive understanding of the user feedback, enabling the ML module 216 to generate accurate summaries at various levels. This sequential flow ensures that the system can process and summarize user feedback effectively, providing the necessary insights for product improvement and feature prioritization.

[0034] The ML module 216 may be a component within the system that transforms user feedback into structured summaries. ML module 216 comprises several processes: identifying key sentences, sentence embeddings, and topic modelling. These processes are designed to work together to extract the essence of user feedback into a concise format. The ML module 216 begins by selecting sentences from the processed user interactions that convey the intended message and sentiment. This selection may be based on the module's algorithms, which assess the relevance of each sentence to the user's feedback. The objective may be to pay attention to the vital information.

[0035] Following sentence selection, the ML module 216 employs embedding techniques to reduce the length of the sentences while preserving the necessary information. This step may be aimed at creating a summary that may be both informative and concise, making it easier to understand the user's feedback. Concurrently, the ML module 216 performs topic modelling to categorize the summaries into groups based on common themes or features. This helps in identifying patterns and insights that are relevant to product development. The ML module 216 may be also engaged in continuous learning, using a dataset of previously processed feedback to enhance its summarization capabilities. This ongoing adaptation allows the module to stay current with changes in language and user expressions, improving the relevance and accuracy of the summaries it generates.

[0036] The input module 212 serves as the gateway for user feedback and queries, initiating the analysis process by capturing diverse user interactions in multiple languages. This input module 212 may be the initial touchpoint for user interactions, necessary for the system's ability to understand and process feedback across various languages. The input module 212 may be designed to capture user feedback and queries in a variety of languages, which may be a foundational step for the multilingual user feedback analysis system 102. This input module 212 may be responsible for interfacing with the user environment, where it collects the raw data necessary for further analysis. The reception of this data occurs under the condition that users provide their feedback or inquiries through the system's supported channels, which could include web forms, emails, or other communication interfaces. Once the data may be received, the input module 212 ensures that it may be in a suitable format for the subsequent processing stages. This involves validating the integrity of the data and confirming that it may be within the expected linguistic parameters that the system can handle. The input module 212 acts as a filter, ensuring that only relevant and processable user interactions are forwarded to the natural language processing NLP module 214.

[0037] The input module 212 operates continuously, as user feedback can arrive at any time, and it must be adept at handling varying volumes of data. It may be the first step in the workflow, setting the stage for the NLP module 214 to perform detailed linguistic analysis. Without the input module's effective operation, the system would be unable to accurately capture and relay user feedback for further processing, which may be essential for generating the actionable insights that drive product improvements and feature prioritization decisions. The NLP module 214 may be a key element in the system, tasked with interpreting and analysing human language. It serves as the bridge between unstructured user feedback and structured insights that drive product improvements. The NLP module 214 consists of several processes that work in tandem to understand and extract meaning from user feedback. These include tokenization, part-of-speech tagging, named entity recognition, and sentiment analysis. Each process plays a specific role in dissecting and interpreting the text to facilitate the generation of accurate summaries. The NLP module 214 begins by tokenizing the received user feedback. This process segments the text into individual words or phrases, enabling the system to analyze the language at a granular level. Tokenization may be essential for understanding the structure of the input and preparing it for further analysis. Part-of-speech tagging may be then applied to each token. Named entity recognition may be employed to detect and categorize key elements within the text, such as product features or user locations. This process identifies these elements and classifies them, which may be necessary for summarizing the feedback in a contextually relevant manner. Sentiment analysis may be conducted to determine the emotional tone behind the user feedback. This step assesses whether the sentiment may be positive, negative, or neutral, providing a qualitative measure of user experience. These processes collectively transform unstructured user feedback into structured data, enabling the ML module 216 to generate concise and informative summaries. The integration of these sub -components ensures that the system accurately captures the essence of user feedback, which may be then used to inform product development decisions.

[0038] The ML module 216 may be a component designed to process user feedback into summaries. It condenses data into a more manageable form, adapting to language variations and changes in user expression patterns. The ML module 216 includes algorithms that analyze user interactions to produce summaries. It identifies sentences that convey the main meaning and sentiment from the feedback, embeds this information, and categorizes feedback by themes. This process aids in the extraction of insights from user feedback, which can be used to inform product development decisions. The ML module 216 takes in processed user interactions that have been analysed by the Natural language processing NLP module 214 to extract key information and sentiment. It then applies sentence compression to shorten the text while preserving necessary information. This may be done by selecting sentences that are most relevant and informative to the feedback's intent.

[0039] Additionally, the ML module 216 uses topic modelling to group related summaries by common themes or features. This organization helps users, such as product managers, to quickly understand trends and prioritize features. To ensure the ML module 216 remains effective, it may be trained on a dataset of previously summarized user feedback and queries, allowing it to adapt to new language patterns and expressions. This continuous learning process ensures the module's summarization capabilities are up to date. The output from the ML module 216 may be formatted into multilevel summaries, providing insights at different levels of detail. These summaries are used to guide product improvements and feature prioritization, supporting decision-making processes within the organization.

[0040] Now referring to figure 3 illustrates a flowchart for multilingual user feedback analysis. The flowchart comprises various steps. Each step may be described below in detail.

[0041] Step 302 encompasses the reception of user interactions by an input module 212 across multiple languages. This step serves as the initial collection point for user-generated data, which may be vital for the subsequent processing stages. The input module 212 may be tasked with capturing feedback and queries from users, ensuring that the system 304 has the necessary raw data for analysis. The input module 212 functions as the interface that collects textual data from various digital platforms where users can submit their feedback or queries. The users provide the feedback, and the system 304 receives and stores it for further processing. The action of receiving entails the capture and storage of user interactions in a format that can be understood and processed by the system 304.

[0042] The purpose of this action may be to collect the initial data from users, which serves as the input for the system 304 to perform analysis and generate insights. This step may be essential for the system 304 to have material to process, and without it, the system 304 would lack the data necessary for summarization and analysis. The input module 212 may be equipped to handle various languages, which requires it to have the capability to accurately capture the linguistic nuances of the user interactions. The input module 212 initiates the process that allows the system 304 to carry out its designated functions as described in the subsequent steps. It may be the first interaction point between the user and the system 304, and its effectiveness has a direct impact on the quality and usability of the data collected for analysis. Step 304 involves the processing of user interactions by a NLP module 214 to extract essential information, sentiment, and identify relevant entities using specific techniques. This step focuses on tokenization, which breaks down text into individual words or phrases. This step may be necessary for the system to analyze the structure of the text and prepare it for further processing. This step includes named entity recognition, where key information within the text, such as names of system features, bugs, or user locations, may be identified and classified. This step aids in summarizing the content at a feature or system level and provides context to the sentiment expressed by the user. The NLP module 214 processes the received user interactions using these techniques to extract meaning and sentiment from the user feedback. This processed information may be then utilized by the ML module 216 to generate informative summaries at various levels, as described in step 306.

[0043] The actions in step 304 and its sub-steps are carried out through the operation of the NLP module 214 within the system 304. The NLP module 214 applies algorithms to perform tokenization, part-of-speech tagging, and named entity recognition on the text data received from the input module 212. These actions are driven by the need to condense user feedback into insights that can inform product development, as outlined in step 310. In summary, step 304 and its sub-steps involve the NLP module 214 executing programmed instructions to process user feedback by breaking down the text into tokens, and recognizing named entities. These actions enable the system 304 to generate accurate summaries that guide product development.

[0044] Step 306 encompasses the process by which the ML module 216 transforms processed user interactions into structured summaries. This step may be divided into sub-steps that detail the specific functions performed by the ML module 216. Sub-step of step 306 involves the ML module 216 analysing the processed data to select sentences or phrases that effectively convey the core meaning and sentiment of the user interactions. The selection may be based on semantic content and context, aiming to identify units of text that will be used to create individual summaries. Further, the ML module 216 applies sentence embedding techniques to the previously identified units of text. This process reduces the length of the text while retaining the necessary information. The goal may be to produce a concise summary that communicates the intended message of the original user interaction. The ML module 216 organizing the summaries by themes or features using topic modelling techniques. This step clusters summaries that relate to similar subjects, facilitating the aggregation of feedback for specific aspects of the system. It also contributes to the generation of daily and monthly summaries by identifying overarching trends and sentiments from the user feedback. The ML module 216 utilizes a dataset comprising previously processed user interactions to train and refine its summarization techniques. This allows the module to adapt to language patterns and expressions found in user feedback. The output from step 306 may be a set of summaries at various levels of granularity, formatted to assist in product development and feature prioritization, aiding in strategic planning.

[0045] Step 308 involves the ML module 216 engaging in an ongoing process of self-improvement. This process may be characterized by the module's ability to adjust its algorithms and improve its performance over time by learning from new data. The processor 202 executes programmed instructions that enable the ML module 216 to analyze processed user interactions, which are stored in the memory 206, and refine its summarization techniques accordingly. Sub-step of 308 specifies that the ML module 216 utilizes a dataset for learning, which includes previously processed user interactions. This dataset serves as a resource for the module to learn from past data. By examining this dataset, the ML module 216 identifies patterns and changes in user feedback. This iterative learning process allows the ML module to update its understanding and enhance the accuracy of the summaries it generates. The continuous learning process described in Step 308 may be essential for the ML module 216 to maintain its performance in the face of evolving language use and user feedback trends. This ongoing adaptation ensures that the summaries produced by the system remain relevant and can be used to guide product development and feature prioritization. The output module 218 then provides these summaries in formats that are useful for informing product improvements and decision-making processes.

[0046] Step 310 involves the output module 218 delivering the processed summaries to users who will utilize this information for product development decisions. The output module 218 formats the summaries into reports or visualizations that are structured to convey the findings and trends from the user feedback analysis. Sub-step of 310 specifies the format in which the output module 218 presents the information, which includes individual summaries, feature-wise summaries, daily system-level summaries, and monthly summaries. Each format addresses different levels of decision-making and analysis. Individual summaries might be used for individual user issues, feature-wise summaries could assist in feature development prioritization, daily summaries provide an overview of feedback trends for the day, and monthly summaries offer a comprehensive view for strategic planning. The output module 218 may be responsible for formatting and presenting the data, driven by the necessity to make the insights derived from user feedback understandable and usable for decision-making. Users who receive the summaries use these insights to guide product development and enhancements, ensuring that the product aligns with user needs and preferences. The effectiveness of step 310 may be measured by the usability, accuracy, and relevance of the information provided, which directly impacts the decisions made based on these insights. The output module 218 must be designed to present the data in a manner that may be interpretable without extensive effort, potentially involving data visualization tools and customizable report features to meet the specific needs of different users.

[0047] Referring now to Figure 4 illustrates a flowchart for feedback analysis, in accordance with an embodiment of the present disclosure.

[0048] The step 402 involves receiving, an input module 212, user interactions in a plurality of languages.

[0049] The step 404 involves processing, by a NLP module 214, the received user interactions to extract essential information and sentiment and identify relevant entities using techniques including breaking down data into individual units and recognizing and classifying important entities in the data.

[0050] Further, the step 406 encompasses the generating, by a ML module 216, condensed information of the processed user interactions at multiple levels, including brief breakdowns of each specific user interaction, grouped insights based on specific features mentioned in the data, overall trends and sentiment related to the entire system captured daily, and broader insights for long-term analysis and trend identification.

[0051] At step 408, involves continuously learning, by the ML module 216, on the processed user interactions, adjusting to changing patterns and expressions to enhance its accuracy in producing embeddings over time.

[0052] At step 410, the system 102 may also be configured for outputting, by an output module 218, the generated multilevel condensed information to inform product improvements and feature prioritization decisions.

[0053] Although implementations for the tiered withdrawal system 102 and the method 400 for multilingual user feedback analysis, have been described in language specific to structural features and methods, it must be understood that the claims are not limited to the specific features or methods described. Rather, the specific features and methods are disclosed as examples of implementations for the system 102 and the method 400 for multilingual user feedback analysis.

Claims

WE CLAIM:

1. A system (102) for analysing user feedback, the system (102) comprising: a memory (206); and a processor (202), wherein the processor (202) is configured to execute programmed instructions stored in the memory (206) for: receiving, an input module (212), user interactions in a plurality of languages; processing, by a natural language processing (NLP) module (214), the received user interactions to extract essential information and sentiment, and identify relevant entities using techniques including breaking down data into individual units, and recognizing and classifying important entities in the data; generating, by a machine learning (ML) summarization module (216), condensed information of the processed user interactions at multiple levels, including brief breakdowns of each specific user interaction, grouped insights based on specific features mentioned in the data, overall trends and sentiment related to the entire system captured daily, and broader insights for long-term analysis and trend identification; continuously learning, by the machine learning (ML) summarization module (216), on the processed user interactions, adjusting to changing patterns and expressions to enhance its accuracy in producing condensed information over time; and outputting, by an output module (218), the generated multilevel condensed information to inform product improvements and feature prioritization decisions.

2. The system (102) of claim 1, wherein the natural language processing (NLP) module (214) processes the received user interactions using tokenization to break down data into individual units.

3. The system (102) of claim 1, wherein the natural language processing (NLP) module (214) processes the received user interactions using part-of- speech tagging.

4. The system (102) of claim 1, wherein the natural language processing (NLP) module (214) processes the received user interactions using named entity recognition to recognize and classify important entities in the data.

5. The system (102) of claim 1, wherein the machine learning (ML) summarization module (216) generates condensed information by identifying key units that best convey the core meaning and sentiment of the data.

6. The system (102) of claim 1, wherein the machine learning (ML) summarization module (216) generates condensed information by condensing selected units while preserving essential information.

7. The system (102) of claim 1, wherein the machine learning (ML) summarization module (216) generates condensed information by grouping related summaries by common themes or features.

8. The system (102) of claim 1, wherein the machine learning (ML) summarization module (216) continuously learns on the processed user interactions using a dataset comprising previously processed data.

9. The system (102) of claim 1, wherein the output module (218) outputs the generated multilevel condensed information in a format comprising brief breakdowns of each specific user interaction, grouped insights based on specific features mentioned in the data, overall trends and sentiment related to the entire system captured daily, and broader insights for long-term analysis and trend identification.

10. A method for analysing user feedback, the method comprising: receiving, an input module (212), user interactions in a plurality of languages; processing, by a natural language processing (NLP) module (214), the received user interactions to extract essential information and sentiment, and identify relevant entities using techniques including breaking down data into individual units, and recognizing and classifying important entities in the data; generating, by a machine learning (ML) summarization module (216), condensed information of the processed user interactions at multiple levels, including brief breakdowns of each specific user interaction, grouped insights based on specific features mentioned in the data, overall trends and sentiment related to the entire system captured daily, and broader insights for long-term analysis and trend identification; continuously learning, by the machine learning (ML) summarization module (216), on the processed user interactions, adjusting to changing patterns and expressions to enhance its accuracy in producing condensed information over time; andoutputting, by an output module (218), the generated multilevel condensed information to inform product improvements and feature prioritization decisions.

Images