1. Field of the Disclosure
The present disclosure relates to a method for providing digital information based on a text document. Further, the disclosure relates to a data processing apparatus, a computer program and to a non-transitory computer-readable medium adapted to perform the method.
2. Description of Related Art
It is known to use electronic data processing devices in conjunction with a display for displaying text to a user. For example, it is becoming increasingly popular for users to read texts, e.g. books and magazines, on devices called “e-book readers” rather than as printed media. Similarly, it has become popular for users to download texts from the internet using a desktop, laptop or tablet computer and to read the texts on a display device that is in communication with or built into the computer. If interested in obtaining further information related to the text, the users may use for example a search engine to enter queries for more information.
The “background” description provided herein is for the purpose of generally presenting the context of the disclosure. Work of the presently named inventors, to the extent it is described in this background section, as well as aspects of the description which may not otherwise qualify as prior art at the time of filing, are neither expressly or impliedly admitted as prior art against the present invention.
It is an object to provide a method of enhancing a user's reading experience.
This object is solved by a method, a data processing apparatus, a computer program and a non-transitory computer-readable medium according to the independent claims.
A method for providing digital information based on a text document comprises generating, for at least a first passage of the text document, first topic data indicative of a topic of the first passage using a first natural language analysis, performing a first search for the digital information using the first topic data, the first search providing a first search result, linking the first search result to the first passage, and displaying the first search result together with the first passage upon a request.
A data processing apparatus comprises a data receiver adapted to receive a text document, a natural language analyzer adapted to perform a first natural language analysis on a first passage of the text document and adapted to generate first topic data indicative of a of a topic of the first passage, a data processor adapted to perform a first search for digital information using the first topic data, the first search providing a first search result, and adapted to link the first search result to the first passage, and a display adapted to display the first search result together with the first passage upon a request.
A computer program includes computer program instructions which when executed by a computer cause the computer to perform a method, the method comprising generating, for at least a first passage of a text document, first topic data indicative of a topic of the first passage using a first natural language analysis, performing a first search for the digital information using the first topic data, the first search providing a first search result, linking the first search result to the first passage, and displaying the first search result together with the first passage upon a request.
Further, a non-transitory computer-readable medium includes the computer program as described.
The foregoing paragraphs have been provided by way of general introduction, and are not intended to limit the scope of the following claims. The described embodiments, together with further advantages, will be best understood by reference to the following detailed description taken in conjunction with the accompanying drawings. The elements of the drawings are not necessarily to scale relative to each other.
A more complete appreciation of the disclosure and many of the attendant advantages thereof will be readily obtained as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings, wherein:
Referring now to the drawings, where like reference numerals designate identical or corresponding parts throughout the several views,
The text document may be any kind of text, such as a text file, a text document to be read or edited in a text editor, a text document to be displayed in a viewer in a viewing format (e.g. the “portable document format” PDF), a web page, an electronically stored book or magazine (e.g. an “e-book”) or any other kind of electronically stored document. For example, the text document may comprise an ASCII representation of one or more portions or an entirety of the text, a Unicode representation of one or more portions or an entirety of the text, an XML representation of one or more portions or an entirety of the text, an HTML representation of one or more portions or an entirety of the text and/or a binary representation of one or more portions or an entirety of the text.
The text document may be stored or located at a device performing the method, at a local server and/or at remote server accessible via a network. The network may be for example a local network, an intranet or the internet, or any other network such as a network for telecommunications or the like.
From the text document, the first passage may be selected. The first passage may be any segment of the text document, such as a segment including several paragraphs, a single paragraph, several or a single sentence, several or a single phrase and/or several or a single word. The segment may for example comprise sub-segments which need not be consecutively located within the text.
Further, also a plurality of passages may be selected at block 100. For example, the text, i.e. the whole text or a whole displayed part of the text, may be automatically divided into a plurality of passages, e.g. into a first to x-th passage. In this case, it may be possible to select a single passage, i.e. the first passage, and/or a subset of the plurality of passages, the subset including a plurality of selected passages.
The first passage (or correspondingly the subset including the plurality of selected passages) may be expressly selected by the reader, e.g. by marking it with a pointing device, by touching a corresponding area of a touch-sensitive display or by a command. Further, the selecting may happen implicitly, e.g. even without notice of the reader, for example when the reader focuses on, stops at or marks the first passage or a corresponding paragraph.
When selected, the first passage may, but need not be highlighted automatically by the method. For example, the first passage may be displayed with another format, e.g. according to a display parameter predetermined for highlighting. The display parameter may be set by a user using the method, or may be determined automatically. When automatically determined, the display parameter may be set in accordance with a semantic content of the first passage, e.g. in accordance with the first topic as determined by the method. For example, the highlighting may include displaying characters or the background in a pre-set color and/or in a bold type, in italics or the like.
Based on the first passage, in particular based on a language of the text of the first passage and/or of its semantic content, the first natural language analysis is carried out for determining the first topic data.
The first natural language analysis may include a semantic analysis and may be based on any kind of technology from the natural language processing field. For example, subject/object relations between words and respective roles of words in sentences may be analyzed. These roles and/or relations may be assigned to the respective words or terms of the first passage during the first natural language analysis, and may be used for determining the first topic.
Further, words of the first passage may be mapped to an existing ontology in order to identify the first topic. For example, words such as “dog” and “cat” may be mapped to “animals”, which may be used for determining the first topic. This allows finding other relevant documents, e.g. those about birds.
Further, when analyzing words, a frequency of occurrence of the respective words in the first passage may be analyzed. Words that are contained frequently in the first passage, and not or only rarely in the rest of the document, are likely to characterize a local topic upon which the first topic data may depend. Such words may be referred to as local keywords. They may be given a higher weight than other words when determining the first topic data in a weighted manner based on words of the first passage.
Still further, the first natural language analysis may also include further techniques such as part-of-speech tagging, syntax parsing, semantic parsing, sentiment analysis and summarization.
Thus, the first topic data may be defined and described based on words, terms, ontologies, sub-topics and/or the like. The first topic may include search terms that may be used e.g. in a search query. The first topic data may thus include, describe and/or be related to an issue, matter, subject or theme of the first passage.
At block 102, a first search for digital information is performed using the first topic data. The first search provides a first search result.
For the first search, any kind of search strategy may be used. For example, a search engine for searching digital information may be used. The search engine may be a specialized search engine which is particularly adapted for carrying out the method and/or a text based search. Alternatively, also a general-purpose search engine such as a publicly available search engine accessible via the Internet may be used, e.g. by submitting a search query and retrieving the result via the Internet. The first topic including various search terms may be used as a basis for the search, e.g. as the search query. The first topic data may thus be supplied to a search engine e.g. as a set of or logical expression based on search terms.
During the first search, any kinds of digital storages may be searched and/or accessed. For example, local data storages, locally accessible files and/or remotely located data which is accessible via the network through which the text document has been accessed, or via another (further) network such as an intranet, the internet or a telecommunications network may be accessed and/or retrieved.
Any kind of digital information may be searched, accessed and/or retrieved as the first search result. For example, at least one data file or a multitude of data files may be accessed, e.g. further text documents related to the first topic data. However, not only text documents, but any kind of digitally encoded and thus searchable information may be accessed, such as web pages, media data or the like.
At block 104, the first search result is linked to the first passage. The linking may include any kind of link, e.g. a pointer to metadata describing the first search result, and/or a plurality of links pointing directly to or including addresses of several pieces of information. For example, and uniform resource identifier, uniform resource locator and/or uniform resource name may be included. The linking may thus allow an easy retrieval of the digital information included in the first search result when the first passage is accessed.
At optional block 106, which may, but need not be performed, content according to the first search result may be loaded, e.g. to a device carrying out the method. For loading, the linking of block 104 may be used. The loading allows retrieval of the search result, for example as soon as it is identified and e.g. in the background, without notice of the user. The loading may include storing the loaded content at a local storage of the device carrying out the method, from which local storage the content may be retrieved instantaneously, e.g. upon a user's request. The loading may include loading the linked content as a whole, and/or loading only a representative part of the content, e.g. a shortcut or summary. The respective part, e.g. summary may be determined with respect to the first topic. The representative part may then be displayed together with an icon representing the linked information.
At optional block 108, which may, but need not be performed, the first passage is displayed according to the display parameter (used for highlighting upon selection of the first paragraph) and/or to a further display parameter. The display parameter and/or further display parameter may be determined in dependence on at least one of the first topic data, the first search result and a semantic content of the first search result.
E.g. when the linking has taken place, the first passage may be formatted such that the user notices that additional content is available. The first passage may be formatted, e.g. highlighted as a whole, or only important keywords may be marked. The (further) display parameter may describe a font, a color, a background color, a setting of characters in bold face, italics or an underlining, and may depend on a user's pre-stored settings.
The display parameter and/or the further display parameter may be determined according to a content of the first paragraph, according to the first topic, and/or according to the first search result and/or to the related content. For example, technical facts may be formatted other than opinionated material. The highlighting may be selected so as to intensify the reading, e.g. by utilizing an underlying mood or emotion. By the highlighting, an availability of further content is indicated to the user, whose interest is raised.
At block 110, the first search result is displayed together with the first passage. The displaying may take place upon a request e.g. entered by the user while reading the text document.
The request may be entered via an input device of a reading device used for reading the text document. For example, at a personal computer, the request may be entered by a mouse click to the first passage, to a marking or to an icon representing the linked and/or loaded first search result. Alternatively, the request may be entered by pressing a request key, by touching a corresponding area on a touch-sensitive panel used as a display, or by uttering a corresponding speech command into a microphone.
When displaying the first search result, contents of pieces of information forming the first search result (e.g. a whole text, text passages, summaries or shortcuts of search results) and/or address data indicating where the contents of the pieces of information may be accessed (e.g. links to search results) may be displayed, as will be described in more detail below.
The first search result may be displayed together with the first passage, e.g. in a vicinity or neighboring area of a display where the first passage is displayed. For example, a particular window or column for displaying the first search result may be displayed near the first passage and/or overlapping the first passage. Thus, the digital information found with respect to the first passage is presented to the reader in conjunction with the first passage, providing an easy and direct access to the digital information. The first search result may also be displayed alone, e.g. in an area in which the text document has been displayed before, the area covering for example the whole display or a whole window in which the text document has formerly been displayed. This kind of displaying may be performed for example if the first search result has been (expressly) selected/requested for display.
Alternatively, the displaying may also take place without a request, e.g. once the first search result is retrieved. In this case, the first search result may be displayed in a separate area not overlapping the display of the text document for not disturbing the user while reading.
It should be noted that the determining of the first passage, the generating of the first topic data, the first search and the linking of the first search result or any combination of steps thereof may be performed in the background. For example, the determination of the first passage may happen without notice of the reader, e.g. when the reader stops at the first passage or at a corresponding paragraph, takes a long time for reading the first passage or focuses thereon e.g. using a pointing device pointing to a corresponding location on the display. The first natural language analysis, first search and/or information retrieval of the first search result may then be performed in the background. The first search result may be presented to the reader without a request or instantly when the user enters a corresponding request, e.g. according to a users pre-stored settings.
Thus, access to the digital information may be provided without an active request, a search, an entering of search terms or the like. The processing may conveniently be performed in the background while the user is reading. Thus, no time for waiting is necessary. In other words, supplemental digital information to the text document may be offered to the reader in conjunction with the text passage he is reading without user interaction and without waiting, as a background service of a text reading software or text reading device.
In
An introductory paragraph 201 is identified as the first passage at 202, e.g. by a mouse click of the user, or by the time the user takes for reading, or the displayed text document is automatically divided in the first to x-th passages.
In the first natural language analysis, keywords of introductory paragraph 210 are identified as important in the first passage, e.g. by analyzing their role as subject or object in sentences, by analyzing their relation, by mapping their content to an existing ontology or by determining their frequency of occurrence in introductory paragraph 210 (first passage) with respect to their frequency in the whole text document or in general language documents. These keywords may be marked e.g. by underlining, as shown.
Based on these keywords, general information on apple cake is determined as the first topic at 204.
On this basis, the first search is performed at 206, using search terms 208 as basis for the search, e.g. as a search filter.
At 210, several pieces of digital information are linked to the first passage as the first search result. The linking is illustrated by the dotted arrow to introductory paragraph 210. In the case depicted, the first search result includes a first link to an intranet page providing an offer of a bistro, e.g. an in-house coffee shop. Further, links to an internet page providing a lexicon and to an intranet page including a database comprising recipes are included. Further, a link to a local data file including personal recipes is provided.
a illustrates effects achieved by the method of
b illustrates effects achieved by the method of
In
At block 400, a second text document included in the first search result may be selected. The second text document may be any document included in the first search result, in particular a document or file including text. At block 402, for at least a second passage of the second text document, second topic data indicative of a topic of the second passage is generated using a second natural language analysis. At block 404, a second search for further digital information may be performed using the second topic data, thereby providing a second search result. At block 406, the second search result may be linked to the second passage.
The second text document may be selected by any kind of selection, e.g. by focusing on it or pointing to it using a pointing device or pointing capabilities of a touch screen. The method steps described in blocks 402 to 406 are then carried out with respect to the second text document. These steps may be performed in a corresponding manner as the method steps illustrated in
The method steps as illustrated in
This results in a kind of text integrated search engine where the user can surf directly from one text document and/or internet page to another based on search engine results and the local topic and the search/browse history. It combines surfing and searching and enables a new mode of accessing the web. The beginning and end of a corresponding search session is given naturally. While a user is looking for a certain topic the user will move from page to page using the above method. When looking for a different topic, the user may however return to a search engine, e.g. for entering a new search query.
In other words, both first and second topic data have influence on a search term 600, in which the terms “apple cake”, “popular” and “traditional” from the first search illustrated in
In the embodiment, the search history from a current session is thus used as an additional resource in each search query. The session history is reflected in each of the iterated search steps. This allows a precise search reflecting the user's interests as known from the current session.
In a further embodiment, it is possible that topic data from a predefined number of (or from all) previous searches is used when performing the first search. Thus, in a search query of e.g. the first search, search terms from the previous searches may be included. Thus, the search history from the current session is used in each iterated search step. Likewise, also a search history from previous sessions may also be included.
When using search history data, a proximity weighting may be applied. This allows forgetting about search topics from the past and concentrating on the user's current interest. However, the search results may be expected to be more precise, e.g. adapted to the users interest, than in single step searches using only search terms which are currently entered by the user.
In the example illustrated in
In
At block 700, a third passage may be selected from the text document. As described for the first passage in the above, the third passage may also include single or multiple words, phrases, sentences or paragraphs, located consecutively or non-consecutively. By using a third natural language analysis, third topic data may be generated at block 702, the third topic data being indicative of a semantic content of the third passage. The third natural language analysis may comprise the same methods as the first and second natural language analysis described in the above. At block 704, a third search may be performed using the first topic data and the third topic data. This provides a refined search result, since in addition to the first topic data, the third topic data is used, e.g. by adding corresponding search terms to the search query. The refined search result is then provided as the first search result at block 706.
The additional and optional method steps of
Effects of these optional method steps are illustrated in the embodiment of
Thus, the first search result may be refined step by step by selection of passages of the text. This enables a convenient navigation towards any digital information of interest, only guided by the user's selected passages and by his reading behavior. It should be noted that again, no manual entering of search terms e.g. to a search bar is necessary. During refinement, also a history-based search and weighting is possible, in which the first topic data may have more or less influence on the third search than the third topic data. For example, a history based proximity weighting of the search terms may be applied.
The search results may be expected to be more precise compared to single step searches with manually entered search queries, since the search is based on the broad context resulting from the selected passages, the whole underlying text and from the session history. Further, the search is convenient since only requiring minimal user interaction and time-saving since most of its parts are carried out in the background.
In another embodiment of the method illustrated in
Corresponding collaborative applications are illustrated in
In a further embodiment illustrated in
An automatic clustering of text document 200 is illustrated in
Accordingly, the first, second and/or third paragraph may be determined by clustering the text document into segments with similar content. The segments may include multiple document paragraphs, sentences, phrases or words, which need not be located consecutively in the text, but are only defined as belonging to the cluster by their topic. Thus, the clustering may be defined by the related topic and/or semantic content, while the document structure may be without influence to the clustering.
In a further embodiment of the method, the first topic data, the second data and/or the third topic data of a respective passage is determined based on the words identified in the respective passage (cluster) and/or in a neighboring passage (cluster). For determining the respective topic data, at least one of the following is used: a frequency of the words in the respective passage, a frequency of the words in the text document, a term frequency-inverse document frequency of the words in the text document and in a set of further text documents, a mapping of the words to an ontology, a respective weight associated with at least one of the words, the respective weight being determined using a frequency of occurrence of at least one word and/or a distance of the at least one word from the selected item, and a semantic analysis of the words.
For example, for determining the first, second and/or third topic data of a respective passage, all stop words in the passage, i.e. words like “and”, “or” or “the” and the like may be ignored. The frequency of the remaining words in the respective passage may be determined. If words occur with a higher frequency, they may be more important for a topic of the passage. Further, words occurring with a high frequency in the respective passage and with a lower frequency in the remaining document may be important for the (local, i.e. first second and/or third) topic of the passage. Such words may be given a higher weight, while words occurring rarely may be given a lower weight. This enables a weighted word-based topic definition. A local topic of a paragraph or section might be refined by differential analysis.
Typically, the “term frequency-inverse document frequency” weight (TF-IDF) may be used for reflecting the importance of a word within the current document with respect to a large global text collection. On this basis, words of importance for topic definition may be identified.
As discussed, words may also be mapped to an ontology, thereby identifying categories of a broader meaning.
A respective weight may be associated with a word based on the distance of the word within the text from a selected item, e.g. a selected text passage marked by the user. A resulting distance-based weighting of word may be useful if no passage is marked, and further if the text is not structured as a sequence of paragraphs.
Further, a semantic analysis of the words reflecting their relations within phrases or sentences, such as their function as subject or object of the sentence, may be performed, thereby identifying the most important words for topic definition.
In a simple embodiment, a list of all words included in the respective passage, but excluding stop words, may be used for topic definition and further for defining the search query. For example, the search query may correspond to the list.
In a further embodiment of the method, the first search result may be displayed in a list upon the request. The list and/or an item of the list may include information representing at least one of a relation between the item and the text document with respect to content, feedback information derived from feedback data regarding the item, the feedback data being previously collected from other users, and a type of content of the item.
For example, the information may be included in the list by displaying items of the list with different colors, fonts or settings. The formatting of the items of the list may depend on a profile or on pre-stored settings of the user.
For example, the information may relate to a relation between the item and the text document with respect to content. For example, it may be visible in the list whether an item relates to more detailed information with respect to the first passage, or whether a corresponding or an alternative opinion is given by the item of the list. Corresponding information may be coded nicely by color or font attributes.
Further, the list may include feedback information. The feedback information may be derived from feedback data regarding the item of the list. The feedback may have been previously collected from other users. For example, a server may note whether users are spending a long time for reading a document, or whether they leave the document quickly. Such feedback may be regarded as implicit feedback. However, the server may also collect explicit feedback provided by the users. If such feedback has been previously collected, it may be reflected in the list. For example, useful items represented in the list may be elevated, while other items may be demoted. Alternatively or in addition, icons or characters reflecting the feedback may be added, such as “+++” for very useful document or “−” for documents of low interest.
Further, the list may be grouped according to a type of content of the items. For example, items relating to lexicon or encyclopedic information may be grouped together, while items relating to content databases, to shopping sites or to social collaboration sites may be grouped separately.
In a further embodiment illustrated in
In another embodiment, when a pointing device is moved on the highlighted text, a related list of keywords and keyword combinations may appear to choose from. This gives the user further control on the topic direction, and allows a convenient focusing on search results.
In a still further embodiment illustrated in
For example, if more information related to the Habsburg Empire is required e.g. by keyword selection while reading text document 200, a shortcut of the best matching document 2300 may be provided. In addition, relevant excerpts from a text body of the best matching document may be presented, and a link to the best matching document may be included. Thus, a relevant part of the best matching text document is displayed upon the request and in conjunction with the first passage. The best match is thus shown without leaving the document. Thus, the user may quickly grasp the best matching information.
The processes, algorithms and electronically driven systems described herein can be implemented via a discrete control device or computing system consistent with the structure shown in
As shown in
The microprocessor 1402 or aspects thereof, in an alternate embodiment, can include or exclusively include a logic device for augmenting or fully implementing this disclosure. Such a logic device includes, but is not limited to, an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), a generic-array of logic (GAL), and their equivalents. The microprocessor can be a separate device or a single processing mechanism. Further, this disclosure can benefit from parallel processing capabilities of a multi-cored CPU.
In another aspect, results of processing or the input of data in accordance with this disclosure, in particular the text document and the first, second and third search result, can be displayed via a display controller 1410 to a monitor 1412. The display controller would then preferably include at least one graphic processing unit for improved computational efficiency. Moreover, the monitor 1412 can be provided with a touch-sensitive interface to a command/instruction interface, and other peripherals can be incorporated, including a scanner, a microphone for entering speech commands or a web cam when image-based data entry is used.
The above-noted components can be coupled to several networks, as shown in
Further, the first, second and third search may access digital information on local and remote storages via network interface 1420 through any of the above networks 1414, 1416, 1418.
Further, the network interface 1420 may provide a communication path to the mobile device, e.g. reading device used for reading the text document and displaying the digital information retrieved. The communication path can be provided by way of packets of data. Alternatively, the displaying of the text document and/or of the digital information retrieved may also take place at monitor 1412, i.e. locally at data processing apparatus 1400.
Further, an I/O interface 1421 may be provided, allowing connecting input devices such as keyboard 1422, mouse 1424 and other peripherals 1426, e.g. a reading device for reading a non-transitory storage medium 1428, such as an optical disk, a magnetic disk, semiconductor memory or the like.
The non-transitory storage medium 1428 may include software, e.g. a computer program including computer program instructions, which when executed by data processing apparatus 1400, causes the computer to perform an embodiment of the above described method. The method may comprise generating, for at least the first passage of the text document, the first topic data indicative of the topic of the first passage using the first natural language analysis, performing the first search for the digital information using the first topic data, the first search providing the first search result, linking the first search result to the first passage, and displaying the first search result together with the first passage upon a request, as discussed in the above.
Additionally, a central BUS 1430 may be provided to connect the above hardware components together and to provide at least one path for digital communication therebetween.
Data processing apparatus 1400 may further comprise a data receiver adapted to receive the text document, which data receiver may e.g. include network interface 1420, disk controller 1406 and I/O interface 1421. Further, a natural language analyzer may be included, which is adapted to perform the first natural language analysis on the first passage of the text document and adapted to generate the first topic data indicative of the topic of the first passage. The natural language analyzer may include inter alia microprocessor 1402 and/or specific parts of memory 1404. Further, a data processor, e.g. including microprocessor 1402, may be included, which data processor may be adapted to perform the first search for the digital information using the first topic data, wherein the first search provides the first search result. The data processor may be adapted to link the first search result to the first passage, e.g. by including addresses of memory 1404 indicating where the first search result and/or corresponding metadata may be stored. Further, data processing apparatus 1400 may comprise a display, e.g. included in monitor 1412 and/or in the mobile device, adapted to display the first search result together with the first passage upon a request.
In an alternate embodiment, the data processing apparatus 1400 may comprise a storage, such as memory 1404, adapted to store the first topic data and topic data from a predefined number of previous searches, wherein the data processor is further adapted to perform the first search using a search query including search terms from the first topic data and the topic data from the predefined number of the previous searches. Thus, an iterated search using history-based search data, search terms and/or search queries may be performed, as described for the various embodiments of the method in the above.
Further, the data processing apparatus 1400 may comprise an input device, e.g. keyboard 1422, mouse 1424, the touch-sensitive interface to the command/instruction interface of monitor 1412, and/or the scanner, the microphone or the web cam for image-based data entry. The input device may be adapted to receive the user selection of the second text document included in the first search result. The second text document may be received by the data receiver, analyzed by the natural language analyzer providing second topic data, and processed in a second search using the second topic data and/or the first topic data, e.g. combined in a weighted manner, by the data processor. A result of the second search may then be displayed at the display, e.g. at monitor 1412 or at the mobile device.
Further, the input device may be adapted to receive a user selection of the third passage from the text document. Then, the third passage may be analyzed by the natural language analyzer providing third topic data and processed in a third search using the third topic data and/or the first topic data, e.g. again combined in a weighted manner, by the data processor. A result of the third search may then be displayed at the display.
Further embodiments of the data processing system are a desktop, laptop or tablet computer (system), an e-book reader or a smart phone.
At reading device 1500, processing capabilities may be provided, including e.g. the first, second and/or third natural language analyzer and the data processor adapted to perform the first, second third search. However, in alternate embodiments of reading device 1500, any parts of these processing capabilities, e.g. the natural language analyzer and/or the data processor performing the first, second and/or third search, may also reside in an external device accessible e.g. via the above-mentioned networks. In this case, e.g. if at least some of the processing steps are performed by a remote, e.g. stationary device and/or a central server, processing capabilities at reading device 1500 may be kept low. The first, second and/or third search result linked to the first, second or third passage may then be stored locally at the reading device and/or remotely at the server device. In the latter case, storage capabilities of reading device 1500 may be kept low. Further, as illustrated in
The above-described devices allow performing a text-integrated search, e.g. a search departing from passages of interest of the text document, which is entirely controlled by the reading interest of the user. When performing the search and departing from the text document, it is not necessary to return to a search bar for entering further search terms. All search terms are drawn from the text document e.g. by the natural language analysis. The search results are expected to be very precise and meeting the users interests, since the search is based on a broad context including the local topic, the whole underlying text of the first text document and/or the session search history.
Obviously, numerous modifications and variations of the present disclosure are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims, the invention may be practiced otherwise than as specifically described herein. Further, there is no limitation in combining any of the above mentioned aspects and features in any way.
Number | Date | Country | Kind |
---|---|---|---|
12005169.3 | Jul 2012 | EP | regional |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/EP13/01669 | 6/6/2013 | WO | 00 |