I Know What You Need: Investigating Document Retrieval Effectiveness with Partial Session Contexts

Reducing user effort in finding relevant information is one of the key objectives of search systems. Existing approaches have been shown to effectively exploit the context from the current search session of users for automatically suggesting queries to reduce their search efforts. However, these approaches do not accomplish the end goal of a search system—that of retrieving a set of potentially relevant documents for the evolving information need during a search session. This article takes the problem of query prediction one step further by investigating the problem of contextual recommendation within a search session. More specifically, given the partial context information of a session in the form of a small number of queries, we investigate how a search system can effectively predict the documents that a user would have been presented with had he continued the search session by submitting subsequent queries. To address the problem, we propose a model of contextual recommendation that seeks to capture the underlying semantics of information need transitions of a current user’s search context. This model leverages information from a number of past interactions of other users with similar interactions from an existing search log. To identify similar interactions, as a novel contribution, we propose an embedding approach that jointly learns representations of both individual query terms and also those of queries (in their entirety) from a search log data by leveraging session-level containment relationships. Our experiments conducted on a large query log, namely the AOL, demonstrate that using a joint embedding of queries and their terms within our proposed framework of document retrieval outperforms a number of text-only and sequence modeling based baselines.

Channel Estimation for Massive MIMO Systems

In MIMO millimeter-wave (mmWave) systems, while the hybrid digital/analog precoding structure provides the ability to increase the reach rate, it also faces the challenge of reducing the channel time limit due to the large number of horns on both sides of the Tx / Rx. . In this paper, channel measurement is done by searching with multiple beams, and a new hierarchical multi-beam search system is proposed, using a pre-designed analog codebook. Performance tests show that, compared to a highperformance system, the proposed system not only achieves a high level of success in getting multiple beams under normal system settings but also significantly reduces channel estimation time Keywords: Massive MIMO, Channel Estimation, precoding

Image Search System Based on Machine Learning

Since the 21st century, with the continuous maturity of network technology and its integration with the education field, traditional face-to-face communication has gradually expanded to the virtual network environment. In the online learning environment, students can use the online platform to communicate directly with teachers, no longer limited by time and region. The time and space breakthrough of teacher-student interaction has brought development opportunities for teachers to constantly contact students with a long-term management mechanism. Based on this situation, this article uses artificial intelligence technology to build a network communication platform. This article first analyzes the application status of artificial intelligence technology in the network communication platform, and then introduces the artificial intelligence technology applied in this article. Then, this article uses artificial intelligence technology to design a network communication platform, and test the function and performance of the platform. The test results show that the function of the system is very accurate and reliable, and the performance of the system is sufficient to support nearly 10,000 users at the same time.

Effect of Sign-recognition Performance on the Usability of Sign-language Dictionary Search

Advances in sign-language recognition technology have enabled researchers to investigate various methods that can assist users in searching for an unfamiliar sign in ASL using sign-recognition technology. Users can generate a query by submitting a video of themselves performing the sign they believe they encountered somewhere and obtain a list of possible matches. However, there is disagreement among developers of such technology on how to report the performance of their systems, and prior research has not examined the relationship between the performance of search technology and users’ subjective judgements for this task. We conducted three studies using a Wizard-of-Oz prototype of a webcam-based ASL dictionary search system to investigate the relationship between the performance of such a system and user judgements. We found that, in addition to the position of the desired word in a list of results, the placement of the desired word above or below the fold and the similarity of the other words in the results list affected users’ judgements of the system. We also found that metrics that incorporate the precision of the overall list correlated better with users’ judgements than did metrics currently reported in prior ASL dictionary research.

The Efficiency of Drones Usage for Safety and Rescue Operations in an Open Area: A Case from Poland

The use of unmanned aerial systems (UAS) is becoming increasingly frequent during search and rescue (SAR) operations conducted to find missing persons. These systems have proven to be particularly useful for operations executed in the wilderness, i.e., in open and mountainous areas. The successful implementation of those systems is possible thanks to the potential offered by unmanned aerial vehicles (UAVs), which help achieve a considerable reduction in operational times and consequently allow a much quicker finding of lost persons. This is crucial to enhance their chances of survival in extreme conditions (withholding hydration, food and medicine, and hypothermia). The paper presents the results of a preliminary assessment of a search and rescue method conducted in an unknown terrain, where groups were coordinated with the use of UAVs and a ground control station (GCS) workstation. The conducted analysis was focused on assessing conditions that would help minimise the time of arrival of the rescue team to the target, which in real conditions could be a missing person identified on aerial images. The results of executed field tests have proven that the time necessary to reach injured persons can be substantially shortened if imaging recorded by UAV is deployed, as it considerably enhances the chance of survival in an emergency situation. The GCS workstation is also one of the crucial components in the search system, which assures image transmission from the UAV to participants of the search operation and radio signal amplification in a difficult terrain. The effectiveness of the search system was tested by comparing the arrival times of teams equipped with GPS and a compass and those not equipped with such equipment. The article also outlined the possibilities of extending the functionality of the search system with the SARUAV module, which was used to find a missing person in Poland.

New Verbal Lexemes of the Word-Formative Family with a Base “Corona” in the German Vocabulary of the Pandemic Period: Structural and Semantic Analysis

The article considers the structural and semantic peculiarities of German colloquial verbs, which belong to the word formation family with the base “Corona”. The article analyzes the lexemes formed during the SARS-CoV-2 pandemic from October 2019 to May 2021. In the course of the study new verbs not listed in the lexicographical sources have been found. The majority of these verbal lexemes are occasional: they are identified by small frequencies and only function within the framework of certain types of text in the Internet discourse (blogs, Twitter, social networks). The author specifies the meaning of some verbs entered in the lexicographical base of the Leibniz-Institute for German Language (OWID.de). The author describes both the formal and the lexical-semantic structures of verbal lexemes, considers their word-forming and lexical-semantic relations that combine the motivating noun “corona” and derived words. The study characterizes the models of building the colloquial verbal lexemes, which are currently productive and highly active in the context of the German “coronavirus discourse”. The study proves that the German vocabulary expanded at the time of Covid-19 pandemic through the suffixal word-forming models and the formation of verbs with prepositional and adverbial particles. The study shows that the models of the formation of verbal units with prepositional and adverbial components are particularly active, while the prefixal models are not active in the formation of verbs with the component “Corona”. The author analyzes examples of the use of the lexemes in context, which are presented in the text corpus of the Google search system, determines the frequency of the verbal units. The article clearly shows the differences in the meaning and functioning of verbal lexemes. The article notes some peculiarities of their lexical compatibility.

The role of historical and contextual knowledge in enterprise search

PurposeThe purpose of the study is to examine enterprise searching practices across different work areas and work tasks in an enterprise search system in an international biotechnology company.Design/methodology/approachA mixed-method approach studying employees' authentic search activities during a 4-month period by log data, questionnaire survey and interviews. The log data analysed the entire active searcher group, whereas the questionnaire and interviews focused on frequent searchers.FindingsThe three studies provided insight into the searching activities and an understanding of the way searchers used the enterprise search system to search for information as part of their work tasks. The data identified three searcher groups, each with specific search characteristics. Four work task types were identified, and for all four types the searchers applied a tracing searching technique with use of contextual and historical relationships as paths.Practical implicationsThe findings point to the importance of knowledge on historical and contextual relations in enterprise search.Originality/valueThe work sheds new light on enterprise searchers' information search practices. A significant contribution is the identification of a tracing search method used in relation to four essential work task types. Another contribution is the importance of historical and contextual knowledge to support the tracing search and decide what paths to follow.

Time Classification Algorithm Based on Windowed-Color Histogram Matching

A web-based search system recommends and gives results such as customized image or video contents using information such as user interests, search time, and place. Time information extracted from images can be used as a important metadata in the web search system. We present an efficient algorithm to classify time period into day, dawn, and night when the input is a single image with a sky region. We employ the Mask R-CNN to extract a sky region. Based on the extracted sky region, reference color histograms are generated, which can be considered as the ground-truth. To compare the histograms effectively, we design the windowed-color histograms (for RGB bands) to compare each time period from the sky region of the reference data with one of the input images. Also, we use a weighting approach to reflect a more separable feature on the windowed-color histogram. With the proposed windowed-color histogram, we verify about 91% of the recognition accuracy in the test data. Compared with the existing deep neural network models, we verify that the proposed algorithm achieves better performance in the test dataset.

ФРАЗЕОЛОГІЧНІ ОДИНИЦІ ТЕМАТИЧНОЇ ГРУПИ “ПОГОДА” В СУЧАСНІЙ НІМЕЦЬКІЙ МОВІ: СТРУКТУРНО-СЕМАНТИЧНИЙ ТА ПЕРЕКЛАДАЦЬКИЙ АСПЕКТИ

The article analyzes the structural and semantic features of German phraseological units of the thematic group “Weather”. This thematic group was chosen for the study given the great importance of weather conditions for various spheres of human life, especially for agriculture, which is undoubtedly refl ected in the language by the presence of a large number of features, descriptions and phraseological units. The study is based on the electronic resource “Oldphras”. Three hundred and four phraseological units, which are the subject of this study, were identifi ed by using the resource search system. The main thematic subgroups are Wetter “weather”, Regen “rain”, Hagel “hail”, Blitz “lightning”, Donner “thunder”, Nebel “fog”, Wind “wind”, Sturm “storm”, Gewitter “bad weather”, Schnee “Snow”, Wolke “cloud”, Frost “frost”, Hitze “heat”, Sonne “sun”, Himmel “sky”, Jahreszeit “season”, Winter “winter”, Frühling “spring”, Sommer “summer”. The electronic resource allows us to accurately understand the meaning of the selected units, as the page has an explanation of each of them, as well as to see whether this phraseology is relevant in modern German and whether it has undergone some changes. The next step was to classify phraseological units according to their structure and semantics. According to the criterion of structure, phraseological units constitute the “phrasicon” of a language – that is, the whole inventory of idioms and phrases, both word-like and sentencelike set expressions. Using these criteria, the fi rst type includes the following compounds: in den Wind reden – “waste (one’s) breath”; Wind haben – “as hungry as a hunter”; in allen Himmeln schweben – “head in the clouds”; Sturm läuten – “to ring the alarm bell”. As for the second type, here are the following examples: Sie hat wohl der Blitz beim letzten Schiß erwischt? – “Are you insane?”; Аhа, daher weht der Wind! – “That’s what the smell is!”; jetzt pfeift der Wind aus einem anderen Loch (jetzt pfeift ein anderer Wind) – “change one`s tune”. The results indicate that sentence-like expressions account for 31.6 % of the entire sample, while word-like ones comprise 68.4 %. Also, we have made use of V. V. Vinogradov’s classifi cation system which is based on the degree of semantic cohesion between the components of a phraseological unit. As a result, the selected phraseological units were classifi ed by translation methods, and it was determined that the most commonly used methods are analogues (41.5 %) and descriptive (36.6 %) ones, while equivalent, combined, antonymous, loan translation, and translation in one word are much less fr).equent (21.9 % altogether). Key words: phraseological unit, translation, semantics, translation equivalence, translation transformations.