GB2589609A

GB2589609A - Recommender system

Info

Publication number: GB2589609A
Application number: GB1917713.8A
Authority: GB
Inventors: Vagliano Iacopo; Nishioka Chifumi; Scherp Ansgar
Original assignee: Ernst and Young GmbH
Current assignee: Ernst and Young GmbH
Priority date: 2019-12-04
Filing date: 2019-12-04
Publication date: 2021-06-09
Also published as: GB201917713D0

Abstract

A recommender system that receives an input at a first timepoint from a user device, the input comprises a unique user identity, and is indicative of a user-intent to initiate a recommendation generation process. A user profile is acquired based on the unique user identity. The user profile includes a set of term-weight data pairs for a number of terms which are derived from a plurality of searches performed at the user device before the first timepoint, and a plurality of user interactions of the user on the user device. A first set of coherent terms is selected based on weight values and a specified set of rules. A search query is generated based on the selected first set of coherent terms and a second set of coherent terms, wherein the second set of coherent terms are derived based on a concept-tree data structure. A set of data items are retrieved in a plurality of different formats from a plurality of different online or offline data sources, and displaying of the retrieved set of data items is such that a visual relevancy of the retrieved set of data items with respect to the first set of coherent terms and the second set of coherent terms is discernible.

Description

RECOMMENDER SYSTEM

TECHNICAL FIELD

The present disclosure relates generally to search platforms and 5 technologies; and more specifically, to recommender systems to provide recommendations of data items, which are of user-interest on user devices.

BACKGROUND

With advancements in computer and communication technologies, and data sharing platforms, there has been rapid increase in the amount of published information. For example, an enormous volume of heterogeneous content, such as scientific publications, videos, tutorials, social media posts, etc. is published almost on daily basis. The amount exceeds by far what users can read in their life time. With the rapid increase in the amount of heterogeneous content and different data sources from which such content can be accessed, users typically struggle to be aware of what knowledge is available which may be relevant to their interest. Currently, many search engines and platforms are available to search for relevant content. However, in certain scenarios, users do not exactly know what to search for, or how to formulate an effective search query. This is more common in explorative tasks such as learning or investigating a new topic. As a result, a large number of documents comprising both relevant and irrelevant information that may be or may not be of user-interest, are typically retrieved and displayed based on the search query, which is time-consuming to visualize, read, and grasp. Moreover, the conventional search platforms and technologies have a diversity problem, i.e., an access to new information or interesting data items not directly related to their current search is restricted or not presented to users. Conventional recommender systems exist, but are still are at nascent stage to accurately gauge user-interest specific to a user and provide recommendations that are diverse but still relevant and of interest specific to the user.

Additionally, as a result of the limitations associated with conventional techniques, the process involved in search query formulation, and retrieval of documents based on search query is resource intensive. For example, retrieval of large amount of information makes the search and retrieval process, significantly processor cycle and memory intensive.

Moreover, irrelevant information occupies unnecessarily high amount of space in a temporary storage device (e.g. Random Access Memory) resulting in unavailability of RAM for performing other tasks, which in turn adversely affects the inherent computational capability of a user device.

Therefore, in the light of foregoing discussion, there exists a need to 15 overcome the aforementioned drawbacks associated with conventional recommender systems, search platforms, and technologies.

SUMMARY

The present disclosure seeks to provide a recommender system. The present disclosure seeks to provide a solution to the existing problem of diversity and retrieval of irrelevant information that is not of user-interest. An aim of the present disclosure is to provide a solution that overcomes at least partially the problems encountered in prior art, and provides recommendation of data items which are diverse and yet relevant in accordance with present interest of a user, and without requiring a search query to be formulated from a given user for a current search.

In an aspect, an embodiment of the present disclosure provides a recommender system comprising: - a processor in a server arrangement configured to: - receive an input at a first timepoint from a user device via a communication network, wherein the input comprises a unique user identity, and is indicative of a user-intent of a user of the user device to initiate a recommendation generation process at the server arrangement; - acquire a user profile of the user of the user device based on the unique user identity in the received input, wherein the user profile comprises a set of term-weight data pairs for a plurality of terms, wherein the plurality of terms are derived from - a plurality of searches performed at the user device before the first timepoint for a specified elapsed time period via a search engine, and - a plurality of user interactions of the user on the user device before the first timepoint, - select a first set of coherent terms from the plurality of terms from the user profile based on at least weight values in the set of term-weight data pairs and a specified set of rules; - generate a search query based on the selected first set of coherent terms and a second set of coherent terms relevant to the selected first set of coherent terms, wherein the second set of coherent terms are derived based on a concept-tree data structure; - retrieve a set of data items in a plurality of different formats from a plurality of different online or offline data sources, based on at least the generated search query; and - control display of the retrieved set of data items as recommendation on the user device such that a visual relevancy of the retrieved set of data items with respect to the first set of coherent terms and the second set of coherent terms is discernible.

Embodiments of the present disclosure substantially eliminate or at least partially address the aforementioned problems in the prior art, and enables retrieval of only relevant information without employing any search query.

Additional aspects, advantages, features and objects of the present disclosure would be made apparent from the drawings and the detailed 5 description of the illustrative embodiments construed in conjunction with the appended claims that follow.

It will be appreciated that features of the present disclosure are susceptible to being combined in various combinations without departing from the scope of the present disclosure as defined by the appended 10 claims.

BRIEF DESCRIPTION OF THE DRAWINGS

The summary above, as well as the following detailed description of illustrative embodiments, is better understood when read in conjunction with the appended drawings. For the purpose of illustrating the present disclosure, exemplary constructions of the disclosure are shown in the drawings. However, the present disclosure is not limited to specific methods and instrumentalities disclosed herein. Moreover, those in the art will understand that the drawings are not to scale. Wherever possible, like elements have been indicated by identical numbers.

zo Embodiments of the present disclosure will now be described, by way of example only, with reference to the following diagrams wherein: FIG. 1 is a block diagram of a recommender system, in accordance with

an embodiment of the present disclosure; and

FIG. 2 illustrate exemplary user interface rendered on a user device displaying a set of data items, in accordance with an

embodiment of the present disclosure.

In the accompanying drawings, an underlined number is employed to represent an item over which the underlined number is positioned or an item to which the underlined number is adjacent. A non-underlined number relates to an item identified by a line linking the non-underlined number to the item. When a number is non-underlined and accompanied by an associated arrow, the non-underlined number is used to identify a general item at which the arrow is pointing.

DETAILED DESCRIPTION OF EMBODIMENTS

The following detailed description illustrates embodiments of the present disclosure and ways in which they can be implemented. Although some modes of carrying out the present disclosure have been disclosed, those skilled in the art would recognize that other embodiments for carrying out or practicing the present disclosure are also possible.

In an aspect, an embodiment of the present disclosure provides a 15 recommender system comprising: -a processor in a server arrangement configured to: - receive an input at a first timepoint from a user device via a communication network, wherein the input comprises a unique user identity, and is indicative of a user-intent of a user of the user device to initiate a recommendation generation process at the server arrangement; - acquire a user profile of the user of the user device based on the unique user identity in the received input, wherein the user profile comprises a set of term-weight data pairs for a plurality of terms, wherein the plurality of terms are derived from - a plurality of searches performed at the user device before the first timepoint for a specified elapsed time period via a search engine, and - a plurality of user interactions of the user on the user device before the first timepoint, - select a first set of coherent terms from the plurality of terms from the user profile based on at least weight values in the set of term-weight data pairs and a specified set of rules; - generate a search query based on the selected first set of coherent terms and a second set of coherent terms relevant to the selected first set of coherent terms, wherein the second set of coherent terms are derived based on a concept-tree data structure; - retrieve a set of data items in a plurality of different formats from a plurality of different online or offline data sources, based on at least the generated search query; and - control display of the retrieved set of data items as recommendation on the user device such that a visual relevancy of the retrieved set of data items with respect to the first set of coherent terms and the second set of coherent terms is discernible.

The aforesaid recommender system provides techniques for efficient and accurate retrieval of data items that are of user-interest as recommendations to a given user. Beneficially, the recommender system uses the acquired user profile to provide recommendations to the given user without the need for the user to provide a search query in the current search. The disclosed recommender system is configured to retrieve data items based on the first set of coherent terms associated with user profile and the second set of coherent terms based on concept-tree data structure, and thus, is able to recommend diverse information which are relevant as well as of user-interest to the given user. Moreover, the display of the retrieved set of data items is clutter free and controlled such that a visual relevancy of the retrieved set of data items with respect to the first set of coherent terms and the second set of coherent terms is discernible. This enables a user to conveniently visualize and quickly grasp the relevant information retrieved from different data sources.

Furthermore, the recommender system is comparatively less computer intensive and requires less storage space as only small chunk of relevant information is occupied in the storage space of a random access memory. Consequently, random access memory is available for performing other tasks of the processor.

The present disclosure provides the recommender system. Throughout the present disclosure, the term "recommender system" refers to a system that is collection of one or more interconnected programmable and/or non-programmable components configured to provide recommendations of the set of data items which are intended to be of user-interest for a given user. Examples include programmable and/or non-programmable components, such as processors, memories, network interface, connectors, and the like. Moreover, the programmable components are configured to store and execute computer instructions for providing the recommendations. In an example, the recommender system provides set of data items such as relevant research documents based on the user-profile of a specific user as recommendations.

The recommender system comprises a processor in the server arrangement. Throughout the present disclosure, the term "processor" relates to a computational element that is configured to respond to and processes instructions that drive the recommender system. Examples of the processor includes, but is not limited to, a microprocessor, a microcontroller, a complex instruction set computing (CISC) microprocessor, a reduced instruction set (RISC) microprocessor, a very long instruction word (VLIW) microprocessor, central processing unit (CPU), a graphical processing unit (GPU), or any other type of processing or control circuitry. Furthermore, the term "processor" may refer to one or more individual processors, processing devices and various elements associated with a processing device that may be shared by other processing devices. Additionally, the one or more individual processors, processing devices and elements are arranged in various architectures for responding to and processing the instructions that drive the recommender system. Throughout the present disclosure, the term "server arrangement" relates to a structure and/or module that include programmable and/or non-programmable components configured to store, process and/or share information. Optionally, the server arrangement includes any arrangement of physical or virtual computational entities capable of enhancing information to perform various computational tasks. Furthermore, it should be appreciated that the server arrangement may be both single hardware server and/or plurality of hardware servers operating in a parallel or distributed architecture. In an example, the server arrangement may include components such as memory, a processor, a network adapter and the like, to store, process and/or share information with other computing components, such as user device/user equipment. Optionally, the server arrangement is implemented as a computer program that provides various services (such as database service) to other devices, modules or apparatus.

The processor in the server arrangement is configured to receive an input at a first timepoint from a user device via a communication network, wherein the input comprises the unique user identity, and is indicative of zo the user-intent of the user of the user device to initiate the recommendation generation process at the server arrangement. Throughout the present disclosure, the term "user device" relates to an electronic device associated with (or operated by) a given user. In other words, the user device enables the user to access and operate the recommender system. Moreover, the user device is intended to be broadly interpreted to include any electronic device that may be used for voice and/or data communication over the communication network. Examples of user device include, but are not limited to, cellular phones, personal digital assistants (PDAs), handheld devices, wireless modems, laptop computers, personal computers, etc. Additionally, the user device includes a casing, a memory, a processor, a network interface card, a microphone, a speaker, a keypad, and a display. It will be appreciated that user refers to any entity including a person (i.e., human being), an organization (i.e. a company, university, and the like), or a virtual personal assistant (an autonomous program or a bot) using the user device and/or system described herein.

The communication network is employed by the user device to transmit the input to the server arrangement. Throughout the present disclosure, the term "communication network" refers to an arrangement of interconnected programmable and/or non-programmable components that are configured to facilitate data communication between one or more electronic devices and/or databases of the user device and processor. Examples of the communication network may include, but is not limited to, one or more peer-to-peer network, a hybrid peer-to-peer network, local area networks (LANs), radio access networks (RANs), metropolitan area networks (MANS), wide area networks (WANs), all or a portion of a public network such as the global computer network known as the Internet, a private network, a cellular network and any other communication system at one or more locations. Additionally, the communication network includes wired or wireless communication that can be carried out via any number of known protocols, including, but not limited to, Internet Protocol (IP), Light Fidelity (Li-Fi), Wireless Access Protocol (WAP), Frame Relay, or Asynchronous Transfer Mode (ATM).

The term "first timepoint" refers to a timestannp of the current input by the user to initiate the recommendation generation process at the server arrangement. Optionally, there is a timestamp associated with each time the user accesses the recommender system. It will be appreciated that unique user identity refers to one or more characteristics of the user which differentiates one user from another user. Examples of unique user identity includes but is not limited to characteristics, a device identity, a unique alphanumeric value assigned to a user or a user profile of the user, a phone number of the user, an email id of the user, and the like. The uniqueness of the user identity enables the processor to identity the user connecting to the recommender system via the user device, so that a correct user profile may be acquired that is specific to the user. It will be appreciated that recommendation generation process refers to a mechanism of the recommender system to generate and provide recommendations to the users accessing the recommender system.

Moreover, the processor is configured to acquire the user profile of the user of the user device based on the unique user identity in the received input, wherein the user profile comprises the set of term-weight data pairs for the plurality of terms. Throughout the present disclosure, the term "user profile" refers to a data log of a given user, wherein the user profile comprises the plurality of terms associated with the given user in a time before the first tinnepoint which describes the preferences of the type of data the user prefers to read and explore. In an embodiment, the processor receives the user profile from the user device via the communication network. In another embodiment, the user profile is created at the server arrangement based on data, such as search history, received from the user device. Optionally, the processor acquires the user profile of the user and stores the user profile in a storage space of the server arrangement. The term "plurality of terms" refers to the words used by the user previously to in prior searches.

The plurality of terms is derived from the plurality of searches performed at the user device before the first timepoint for the specified elapsed time period via the search engine. In an embodiment, the processor in the server arrangement is further configured to determine a time factor for each term of the plurality of terms, wherein the time factor of a term from the plurality of terms is a timestamp of a term last searched at the user device via the recommender system or other search engines.

Notably, the users search for information required by them on different search engines or via a search engine of the recommender system by inputting one or more terms related to the information. The tinnestannp of a given term refers to a data about a time the term was last used by the user for searching. The timestamp of the given term enables the processor to know a time based relevance of the term to the user. In an example, terms employed for searching by the user in the last 5 days may be more relevant to the user compared to the terms employed by the user in the last 6 months. In such an example, tinnestamp of the terms enables the processor to know about the relevancy and latest user-interest of the terms to the user. As mentioned previously, the plurality of searches of the user, prior to the first tinnepoint are employed by the processor. Optionally, the plurality of searches performed by the user before the first tirnepoint is stored in a user specific database, as user profile.

The plurality of terms is derived from the plurality of user interactions of the user on the user device before the first timepoint. It will be appreciated that the plurality of user interactions of the user refers to activities of the user upon receiving results of the plurality of searches, performed by the user prior to the first tinnepoint. Examples of the plurality of user interactions may include but are not limited to a rating given to a document by the user, number of clicks by the user on a specific type of document, number of downloads of a specific type of document by the user, or other activities which indicates user-interest for a certain topic or concept. In such an example, the plurality of terms are derived from the terms used by the user for searching the documents received as results, which were then clicked, rated, or were downloaded by the user. Optionally, the documents downloaded by the user may be given a higher weight compared to the documents rated by the user. Optionally, the documents rated by the user may be given a higher weight compared to the documents clicked by the user.

In an embodiment, the processor in the server arrangement is further configured to determine a hit factor for each term of the plurality of terms, wherein the hit factor of the term of the plurality of terms is a number of times the term is searched before the first tinnepoint by the 5 user with respect to a total number of searches performed by the user at the user device before the first tinnepoint. Optionally, plurality of user interactions is also employed for determining the hit factor of the term. More optionally, terms present in the documents which are rated by the user, clicked by the user, downloaded by the user, are used for 10 determining the hit factor. It will be appreciated that hit factor enables the processor to determine the terms which are relevant to the user.

In an embodiment, the processor in the server arrangement is further configured to generate a weight for each term of the plurality of terms in the set of term-weight data pairs based on the determined time factor and the hit factor for each term of the plurality of terms. It will be appreciated that term-weight data pairs for the plurality of terms refers to a weight associated with each term of the plurality of terms. Specifically, weight refers to a score associated with each term of the plurality of terms. In an example, the weight of a specific term is calculated based on a number of times the specific term is used by the user for searching. In another example, the weight of a specific term is calculated based on a frequency of term usage in recent time by the user. It will be appreciated that a term used by the user in last seven days is more important and thereby highly weighted compared to a term used by the user in last three months.

In an example, the weight of each term of the plurality of terms is calculated by mathematical equation (1).

W=at ft + ah fh (1) wherein 'W' denotes the weight of the term, 'at' denotes the time coefficient, 'Oh' denotes the hit coefficient, 'ft' denotes the time factor and 'fh' denotes the hit factor. The time factor is the tinnestannp (t) of a term's last search, normalized by the current time (T), such that ft = t/T. The hit factor is the number of times the term has been looked up by the user (h) divided by total number of searches made by the user (H), such that fh = h/H. The user profile is the term-weight data pairs such as <ki, wherein ki is a term and wi a weight of the term.

Furthermore, the processor is configured to select the first set of coherent terms from the plurality of terms from the user profile based on at least weight values in the set of term-weight data pairs and a specified set of rules. The term "first set of coherent terms" refers to a the set of N terms which have the highest weight values. The threshold N is based on the specified set of rules. The specified set of rules refers to protocols and/or parameters that have been defined for calculating the threshold N. In an example, the first set of coherent terms may be a single word or a group of related words, which may indicate a single concept. Thus, to find most relevant data items which are also of user-interest, the selection of at least one set of coherent terms from hundreds of terms (i.e. the plurality of terms) may be executed. In an example, the weight value of a first term, a second term, a third term, a fourth term and a fifth term is 10, 20, 15, 5, and 12 respectively. In such an example, the threshold N may be 3 and therefore, the first set of coherent terms includes the second, the third and the fifth term.

Moreover, the processor is configured to generate the search query based on the selected first set of coherent terms and the second set of coherent terms relevant to the selected first set of coherent terms, wherein the second set of coherent terms are derived based on a concept-tree data structure. Throughout the present disclosure, the term "concept-tree data structure" refers to a hierarchical structure of terms wherein each term has relations with one another. Each term in the concept-tree data structure is represented by a node. In such case, each node is connected to one another via a node-link structure. Typically, in the concept-tree data structure, each node includes a pointer (namely, address) to a parent node. It will be appreciated that the node may or may not have a child node. Consequently, the node may or may not include a pointer to the child node. Moreover, the node may have 0, 1, 2, 3, and so on, number of child node associated therewith. Typically, the concept-tree data structure is instigated by a root node (namely, the starting point of the tree), wherein the root node is the highest-level node. The concept-tree data structure is terminated by leaf nodes (namely, the ending point of the tree), wherein the leaf nodes are the bottom-level nodes. It will be appreciated that the second set of coherent terms are derived based on the relations that the first set of coherent terms in the concept-tree data structure have with other terms in the concept-tree data structure. In an example, the second set of coherent terms are the terms in the concept-tree data structure which are closely related to the first set of coherent terms. In another example, for a node in the concept-tree data structure is a broader term to the leaf node and a narrower term to the root node. The term "search query" herein refers to a string of characters which collectively represent and relate to the first set of coherent terms and the second set of coherent terms.

In an example, if a user profile includes the first set of coherent terms (i.e. a concept) named "Open innovation" and assuming the latter is a sub concept of "Innovation management" which in turn is a sub concept of Management, then the processor may be configured to assign nonzero weights to the concepts "Innovation management" and "Management", even if they are not directly mentioned in user profile or a document. In this way, if the first set of coherent terms, "Innovation management" is part of the user profile, then the processor maybe configured to identify second set of coherent terms, e.g. "Open innovation" automatically, also the data items related to "Open innovation" may be recommended. Similarly, if the first set of coherent terms, i.e., "Open innovation" is part of the user profile, then also the data items related to the concept "Innovation management" may be suggested. This enables the recommender system to generate more diverse recommendations since diverse data items not directly related to the user profile but still relevant to it are considered.

In an embodiment, the processor in the server arrangement is further configured to derive the second set of coherent terms from the selected first set of coherent terms based on a spreading-activation function applied on the selected first set of coherent terms and a machine-readable thesaurus. The spreading-activation function is configured to derive the second set of coherent terms by assigning weights to the terms connected to the first set of coherent terms in the concept-tree data structure. In an example, the first set of coherent terms includes a term such as 'open innovation'. In such an example, the second set of coherent terms includes terms such as 'innovation management' and 'management'. The second set of coherent terms enables in generation of more diverse recommendations, as described in above example. The machine-readable thesaurus comprises a collection of terms arranged in a way that each term is linked to another term such that the linked terms are similar in meanings. The machine-readable thesaurus further improves the diverse recommendation. Optionally, the processor is configured to derive the second set of coherent terms from the machine-readable thesaurus without any human intervention.

In an example, the Bell Log spreading-activation function is used to derive the second set of coherent terms. The Bell Log spreading-activation function is provided by equation (2): BL(c,d)= nih 1 EckjEdj nkj + log10(nodes(h(c)+1)).E BL(ci,d) (2) wherein, h(c) refers to a level where a term 'c' is located in the concept-tree data structure, while nodes counts the concepts at a given level in the tree. CL is a set of terms located in one level lower than the term 'Ci considered.

In an example, the weight is assigned to the second set of terms based on the equation (3) VV, = BL (c d) log 0.1,9 el dB (3) wherein, Ws refers to weight of the second set of term, D is the set of data items which can be recommended and {d: c, e d} is the set of data items which contains the concept c,.

Furthermore, the processor is configured to retrieve the set of data items in a plurality of different formats from a plurality of different online or offline data sources, based on at least the generated search query. In an embodiment, the set of data items in the plurality of different formats comprises at least two of: a text-containing document, a video, an audio, an image file, a social media item, a web-page, or another document format. Each of the different formats of the set of data items comprise information which is useful for the user and thereby is recommended to the user. In an example, the information in a video is identified based on tagged metadata of the video, indexed text of audio of the video. In another example, the information in an audio is identified based on metadata of the audio, or by audio-to-text conversion. In yet another example, the information in an image is identified based on metadata of the image, and image-to-text conversion for textual content in the image. In another example, the information in a social media item and web-page is identified based on respective metadata, and a crawled and indexed text of the social media item and web-page. Optionally, the data source may be a database, a hardware, software, firmware and/or any combination thereof. The term "set of data items" refers to information that is recommended to the user by the system. The different formats of the set of data items include but are not limited to.pdf, .doc, .docx, .jpeg, .png, .xls, .html, .txt, .nnp4, .mp3 and the like. It will be appreciated that the offline data sources are pre-stored in a database of the recommender system In an embodiment, the processor in the server arrangement is further configured to retrieve the set of data items based on parsing of a title or a combination of the title or an abstract of a data item of the set of data items, in absence of full-text of the data item. Optionally, the full text of a given data item may not be readily available for various reasons such as a need for user login for accessing full text, a need for payment for accessing full text and the like. However, in such a case, the title and/or the abstract of the given data item is readily available. Moreover, a given user may read the title and/or abstract of the given data item and access the full text by visiting a web link provided by the processor. In an example, transcripts of a video may not be available to enable the processor to retrieve the transcripts.

In an embodiment, the plurality of different online data sources corresponds to a plurality of web-based platforms from which data items are retrieved on-the-fly, and wherein the plurality of offline data sources corresponds to pre-stored data items in the plurality of different formats in the recommender system. In an example, the web-based platforms may include research paper based platforms, academic information platform of universities and the like. Optionally, the recommender system is configured to store the plurality of data sets available on the plurality of web-based platforms in a downloaded form in the offline data sources.

It will be appreciated that the set of data items are retrieved based on the generated search query. In an example, the search query is used to match with the information available in form of text-containing documents, videos, audios, image files, social media items, web-pages, or another document format. Moreover, the information matching the search query retrieved by the processor.

Furthermore, the processor is configured to control display of the retrieved set of data items as recommendation on the user device such that the visual relevancy of the retrieved set of data items with respect to the first set of coherent terms and the second set of coherent terms is discernible. The processor is configured to employ a display module to display the retrieved set of data items as recommendation. The term "display module" refers to a structure including an arrangement of interconnected programmable and/or non-programmable components that are configured to receive set of data items and present the set of data items on a display screen. Moreover, the display module has a circuitry comprising one or more electronic components which enables the display module to display the set of data items. Optionally, the display module may display the first set of coherent terms. More optionally, the display module may display the second set of coherent terms. More optionally, the display module may display both the first set of coherent terms and the second set of coherent terms to select certain concepts for filtering of search results. It will be appreciated that a user interface on the display module enables the user to interact (such as click, read) with the set of data items. Optionally, the user interface comprises a separate section for providing recommendations. More, optionally, the user interface comprises a first section to enable the user to search for information by entering a search query and a second section to provide recommendations to the user. In an example, the user interface comprises a user interface element to display the recommendations. In such an example, the user interface also comprises a second user interface element enabling the user to perform a plurality of searches.

The visual relevancy of the retrieved set of data items refers to displaying of the set of data items such that a relevancy of the displayed search results and user-interest indicated by first set of coherent terms (i.e. first concept) and the second set of coherent terms (i.e. second concept or sub-concept), may be easily discernible and understood from the displayed view.

DETAILED DESCRIPTION OF THE DRAWINGS

Referring to FIG. 1, illustrated is a block diagram of a recommender system 100, in accordance with an embodiment of the present disclosure. As shown, the recommender system 100 comprises processor 102 in a server arrangement 104. The server arrangement 104 is connected to a user device 106 via a communication network.

Referring to FIG. 2, illustrated is an exemplary user interface 200 rendered on a user device displaying a set of data items, in accordance with an embodiment of the present disclosure. As shown, the user interface 200 comprises a first user interface section 202 displaying the set of data items recommended to a user. The first user interface section 202 comprises a first data item 204 displaying a video. Moreover, the first user interface section 202 comprises a second data item 206 displaying a document. Furthermore, the first user interface section 202 comprises a third data item 208 displaying a web-page. As shown, the user interface 200 comprises a second user interface section 210 which enables the user to perform a plurality of searches.

Modifications to embodiments of the present disclosure described in the foregoing are possible without departing from the scope of the present disclosure as defined by the accompanying claims. Expressions such as "including", "comprising", "incorporating", "have", "is" used to describe and claim the present disclosure are intended to be construed in a nonexclusive manner, namely allowing for items, components or elements not explicitly described also to be present. Reference to the singular is also to be construed to relate to the plural.

Claims

CLAIMSWhat is claimed is: 1. A recommender system comprising: -a processor in a server arrangement configured to: -receive an input at a first timepoint from a user device via a communication network, wherein the input comprises a unique user identity, and is indicative of a user-intent of a user of the user device to initiate a recommendation generation process at the server arrangement; -acquire a user profile of the user of the user device based on the unique user identity in the received input, wherein the user profile comprises a set of term-weight data pairs for a plurality of terms, wherein the plurality of terms are derived from: - a plurality of searches performed at the user device before the first timepoint for a specified elapsed time period via a search engine, and - a plurality of user interactions of the user on the user device before the first timepoint, - select a first set of coherent terms from the plurality of terms from the user profile based on at least weight values in the set of term-weight data pairs and a specified set of rules; - generate a search query based on the selected first set of coherent terms and a second set of coherent terms relevant to the selected first set of coherent terms, wherein the second set of coherent terms are derived based on a concept-tree data structure; - retrieve a set of data items in a plurality of different formats from a plurality of different online or offline data sources, based on at least the generated search query; and - control display of the retrieved set of data items as recommendation on the user device such that a visual relevancy of the retrieved set of data items with respect to the first set of coherent terms and the second set of coherent terms is discernible.
2. The recommender system according to claim 1, wherein the processor in the server arrangement is further configured to determine a time factor 5 and a hit factor for each term of the plurality of terms, wherein the time factor of a term from the plurality of terms is a timestamp of a term last searched at the user device via the recommender system or other search engines, and wherein the hit factor of the term of the plurality of terms is a number of times the term is searched before the first tinnepoint by 10 the user with respect to a total number of searches performed by the user at the user device before the first timepoint.
3. The recommender system according to claim 2, wherein the processor in the server arrangement is further configured to generate a weight for each term of the plurality of terms in the set of term-weight data pairs based on the determined time factor and the hit factor for each term of the plurality of terms.
4. The recommender system according to claim 1, wherein the set of data items in the plurality of different formats comprises at least two of: a text-containing document, a video, an audio, an image file, a social media 20 item, a web-page, or another document format.
5. The recommender system according to claim 1, wherein the plurality of different online data sources corresponds to a plurality of web-based platforms from which data items are retrieved on-the-fly, and wherein the plurality of offline data sources corresponds to pre-stored data items in the plurality of different formats in the recommender system.
6. The recommender system according to claim 1, wherein the processor in the server arrangement is further configured to derive the second set of coherent terms from the selected first set of coherent terms based on a spreading-activation function applied on the selected first set of coherent terms and a machine-readable thesaurus.
7. The recommender system according to claim 1, wherein the processor in the server arrangement is further configured to retrieve the set of data items based on parsing of a title or a combination of the title or an abstract of a data item of the set of data items, in absence of full-text of the data item.